A Novel Markov Decision Process-Based Solution for Improved Quality Prioritized Video Delivery

The recent growth in both number of high specification mobile devices and network multimedia data demand, has made difficult supporting access to the multimedia content at high user Quality of Experience (QoE) levels. It is even more challenging to support offering such services adjusted to user specific requirements in current heterogeneous wireless network environments (HWNE). This paper proposes a novel Prioritized Adaptive Real-time Multi-user Access Network Selection framework (P-ARMANS) which employs a Markov Decision Process (MDP) solution to perform improved bandwidth resource allocation. P-ARMANS enables load balancing during multimedia delivery when users with diverse priority and different device screen resolutions access various services. Modeling and simulations show how P-ARMANS distributes content with different types of traffic and load among typical and business users at higher QoE levels compared to a classic no-priority approach.